Millions of people may have been left vulnerable to hackers while surfing the web on Apple and Google devices, thanks to a newly discovered security flaw known as “FREAK attack.”
There’s no evidence so far that any hackers have exploited the weakness, which companies are now moving to repair. Researchers blame the problem on an old government policy, abandoned over a decade ago, which required U.S. software makers to use weaker security in encryption programs sold overseas due to national security concerns.
Many popular websites and some Internet browsers continued to accept the weaker software, or can be tricked into using it, according to experts at several research institutions who reported their findings Tuesday. They said that could make it easier for hackers to break the encryption that’s supposed to prevent digital eavesdropping when a visitor types sensitive information into a website.
About a third of all encrypted websites were vulnerable as of Tuesday, including sites operated by American Express, Groupon, Kohl’s, Marriott and some government agencies, the researchers said. University of Michigan computer scientist Zakir Durumeric said the vulnerability affects Apple web browsers and the browser built into Google’s Android software, but not Google’s Chrome browser or current browsers from Microsoft or Firefox-maker Mozilla.
From the researchers: “Weak algorithms”
The researchers writing at https://www.smacktls.com/#freak explain how the vulnerability happened:
This attack targets a class of deliberately weak export cipher suites. As the name implies, this class of algorithms were introduced under the pressure of US governments agencies to ensure that they would be able to decrypt all foreign encrypted communication, while stronger algorithms were be banned from export (as they were classified as weapons of war).
Support for these weak algorithms has remained in many implementations such as OpenSSL, even though they are typically disabled by default; however, we discovered that several implementations incorrectly allow the message sequence of export ciphersuites to be used even if a non-export ciphersuite was negotiated.
Thus, if a server is willing to negotiate an export ciphersuite, a man-in-the-middle may trick a browser (which normally doesn’t allow it) to use a weak export key. By design, export RSA moduli must be less than 512 bits long; hence, they can be factored in less than 12 hours for $50 on Amazon EC2.
Ironically, many US government agencies (including the NSA and FBI), as well as a number of popular websites (IBM, or Symantec) enable export ciphersuites on their server – by factoring ther 512-bit RSA modulus, we can impersonate them to vulnerable clients.
Apple Inc. and Google Inc. both said Tuesday they have created software updates to fix the “FREAK attack” flaw, which derives its name from an acronym of technical terms.Apple said its fix will be available next week and Google said it has provided an update to device makers and wireless carriers.
A number of commercial website operators are also taking corrective action after being notified privately in recent weeks, said Matthew Green, a computer security researcher at Johns Hopkins University.
But some experts said the problem shows the danger of government policies that require any weakening of encryption code, even to help fight crime or threats to national security. They warned those policies could inadvertently provide access to hackers.
“This was a policy decision made 20 years ago and it’s now coming back to bite us,” said Edward Felten, a professor of computer science and public affairs at Princeton, referring to the old restrictions on exporting encryption code.